USING ADMINISTRATIVE AND CLINICAL DATA TO DETECT DRUG USE AND HIV RISK IN FOSTER CARE Project Summary The candidate's goal is to develop an independent research career as a developmental psychologist focused on identifying mechanisms to prevent the onset of pervasive and detrimental health risk behaviors, including substance abuse and HIV risk behaviors, among adolescents and young adults involved in foster care. She has experience in examining psychosocial and behavioral processes during this period of development and a strong publication record for the current phase in her career. Her previous training has emphasized development during adolescence and the transition to adulthood and health behaviors during this timeframe. In the K01, she proposes to expand her training to become an expert leader in team science approaches to address substance use and HIV risk behaviors during adolescence and the transition to adulthood, specifically focusing on foster youth, using child welfare administrative data and electronic medical records (EMR). Her training will emphasize: 1) development as an independent researcher and leader in interdisciplinary approaches to addressing health and development among foster youth, 2) substantive training in child maltreatment, substance use, and HIV risk behaviors, and 3) the use of machine learning to distinguish youth at risk, with the end goal of identifying opportunities for prevention. This will be accomplished through seminars and journal clubs offered at Cincinnati Children's Hospital Medical Center (CCHMC), attending regional and national conferences and trainings on substantive topics, and completion of coursework in child maltreatment and machine learning. Finally, she will complete coursework and seminars addressing the ethical conduct of research. Dr. Beal's goal is to submit an R01, as PI, during the third year of this grant. CCHMC, one of the top three pediatric research hospitals in the nation, is committed to Dr. Beal's career development. The research environment at CCHMC and within the Division of Behavioral Medicine and Clinical Psychology, and the mentors and collaborators identified in this proposal, are well-suited to Dr. Beal's interests and research goals. Dr. Beal will be co-mentored by Dr. Jennie Noll and Dr. Robert Ammerman. Dr. Noll has extensive experience as a researcher examining the impact of child maltreatment on adolescent development, particularly around sexual risk behaviors. Dr. Ammerman is a well-established researcher with expertise in prevention and intervention programs for high-risk populations. Her advisory committee includes Dr. Melissa Jonson-Reid, an internationally recognized researcher in child welfare and the use of administrative records, Dr. John Schulenberg, developmental psychologist and authority in adolescent and young adult substance use, Dr. Jessica Kahn, expert in adolescent medicine and sexual and HIV risk behaviors, and Dr. John Pestian, leader in machine learning and its application to clinical care. These individuals have made a commitment to mentor and guide Dr. Beal in research and career development. The objective of this research program is to fill a gap in our current understanding of which adolescents in foster care are at greatest risk for substance use and HIV while simultaneously providing a means for establishing Dr. Beal's independence as a researcher. There are two aims of this proposal. Aim 1 uses discrete time survival analyses and linked child welfare and EMR data to examine the onset and prevalence of substance use and HIV risk behaviors for youth ages 10-20 in foster care compared to matched Medicaid-eligible youth not in foster care (Foster care n=2836; 1-to-1 matched comparison from ~9585 primary care patients). Findings will provide estimates of the rates of substance use and HIV risk behaviors across ages 10-20 and when they increase/decrease, which is critical for identifying the ideal time to begin prevention and intervention efforts. Aim 2 will draw on causal inference techniques and machine learning to identify youth who are substance using or at substantial risk for developing HIV prior to age 21 based on two categories of risk factors: 1 ? those common in adolescence, including parental substance use and mental health diagnoses; and 2 ? risk factors unique to foster care, including extended time in foster care and placement instability. Analyses will be conducted using structural equation modeling to identify predictors of substance use and HIV risk behaviors. Machine learning will be used to develop classifiers that distinguish youth likely to use substances and contract HIV. This approach will, for the first time, provide a means to identify youth at risk for these negative health outcomes before youth experience these consequences. Via this approach, clinicians and front-line workers will be able to easily distinguish youth at risk and dedicate more time to prevention efforts. The proposed program of training and research will provide a clear pathway for Dr. Beal to transition as an independent researcher. This unique program of research merges her expertise in developmental science and health during adolescence and the transition to adulthood with research in child maltreatment, substance use, and HIV risk behaviors to support interdisciplinary, innovative team science approaches to prevent poor health outcomes among foster youth. The proposed studies in this application are significant because the findings take a developmental perspective to provide clinically useful information about which youth are at risk for poor health outcomes. Such an approach is innovative and will transform prevention delivery efforts for vulnerable foster youth.